Article Text

Download PDFPDF

Original research
Bilateral inferior petrosal sinus sampling using desmopressin or corticotropic-releasing hormone: a single-center experience
  1. Amy R Deipolyi1,2,
  2. Bailin Alexander1,2,
  3. Jun Rho1,2,
  4. Joshua A Hirsch2,3,
  5. Rahmi Oklu1,2
  1. 1Division of Interventional Radiology, Massachusetts General Hospital, Boston, Massachusetts, USA
  2. 2Harvard Medical School, Boston, Massachusetts, USA
  3. 3Division of Interventional Neuroradiology, Massachusetts General Hospital, Boston, Massachusetts, USA
  1. Correspondence to Dr Rahmi Oklu, Massachusetts General Hospital, 55 Fruit Street, GRB 290A, Boston, MA 02114, USA; roklu{at}


Background Bilateral inferior petrosal sinus sampling (BIPSS) following corticotropic-releasing hormone (CRH) stimulation is the current gold standard technique in the diagnosis of Cushing disease. However, as a result of CRH shortage, desmopressin (DDAVP) has been used instead for BIPSS. We present the experience of a single tertiary care center using the modified BIPSS protocol and compare the results obtained with DDAVP with those obtained with CRH.

Methods Using the radiology department's electronic database, BIPSS procedures performed at our institution using DDAVP and CRH were identified. Electronic medical records and imaging studies were reviewed and the clinical history, demographic data, endocrine test results, complications of BIPSS, and patient outcomes were recorded. BIPSS data were analyzed for centralization and lateralization of pituitary adrenocorticotropic hormone (ACTH) source. We identified 20 BIPSS cases (16 women, mean age 38 years) performed using DDAVP between 2012 and 2013.

Results The 20 cases demonstrated conventional inferior petrosal sinus anatomy and were successfully cannulated bilaterally. Of these, 18 met the criteria for both centralization and lateralization. A total of 18 patients underwent trans-sphenoidal tumor resection; one patient was lost to follow-up and one is still being followed. There were no complications resulting from the use of DDAVP, specifically no thromboembolic events. Calculated sensitivity for BIPSS with DDAVP was 94.5%. There was also no significant difference in the biochemical results produced by BIPSS using either DDAVP or CRH. All 18 patients demonstrated an ACTH-secreting adenoma on pathology review.

Conclusions DDAVP is a safe alternative to CRH, producing comparable diagnostic results.

  • Angiography
  • Vein
  • MRI
  • Tumor

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Cushing syndrome (CS) is an endocrine disorder caused by prolonged hypercortisolism resulting in osteoporosis, glucose intolerance, hypertension, hypercoagulable state, and neuropsychiatric symptoms. Establishing the cause of CS, which has an incidence of approximately one per million, is essential in management given its considerable morbidity and increased mortality. Most endogenous CS is adrenocorticotropic hormone (ACTH)-dependent, usually arising from ACTH-producing pituitary gland adenomas, a condition referred to as Cushing disease (CD). Bilateral inferior petrosal sinus sampling (BIPSS) is the gold standard technique to differentiate CD from ectopic ACTH production, and allows for minimally invasive measurement of ACTH production by the pituitary gland.1 ,2 When performed in conjunction with corticotropic-releasing hormone (CRH) stimulation, the diagnostic accuracy of BIPSS exceeds 97%.1 ,2 BIPSS is usually reserved for cases where the diagnosis remains elusive prior to definitive surgical management.

However, a recent shortage of CRH led to the use of desmopressin (DDAVP), a synthetic form of vasopressin which can also stimulate pituitary ACTH secretion, but only limited data are available on its use in BIPSS. Some have suggested that, because patients with CS are characterized by a hypercoagulable state, DDAVP may not be safe as it can increase levels of von Willebrand factor (vWF). Theoretically, this could cause platelet adhesion and consequently thromboembolism.3 Indeed, thromboembolism is a rare complication of BIPSS that occurs when BIPSS is performed without prophylactic intraprocedural heparin.4 ,5 However, in addition to vWF, tissue plasminogen activator levels also rise in response to DDAVP, possibly inducing a hematologic balance by activating fibrinolytic pathways.6 Thus, the increased risk of thromboembolic events in patients with CS using DDAVP is probably a theoretical consideration. However, given that this is a grave complication, we sought to evaluate the safety and efficacy of using DDAVP at our institution.

We report a single-center series of 20 patients who underwent BIPSS with DDAVP to evaluate its safety and diagnostic accuracy in comparison with previously performed BIPSS using CRH stimulation.

Materials and methods

This IRB-approved study included all patients who underwent BIPSS with DDAVP stimulation between April 2012 and February 2013. All patients had presumed ACTH-dependent CS based on one or more laboratory results such as 24 h urine free cortisol level, late night salivary cortisol level, plasma ACTH level, dexamethasone suppression test, and imaging studies.

Our institutional protocol for BIPSS with DDAVP is similar to the BIPSS protocol with CRH stimulation, which involves positioning catheters in both inferior petrosal sinuses and then collecting baseline samples from a peripheral line and from both sinuses (figure 1).2 Following peripheral intravenous infusion of 10 µg DDAVP in normal saline, samples are collected at 3, 5, 10 and 15 min. The diagnostic criteria for CD by BIPSS with DDAVP include the same criteria as are used for CRH stimulation (ie, inferior petrosal sinus/peripheral (IPS/P) ACTH ratios of ≥2 at baseline or ≥3 after stimulation).7 Lateralization is defined as an ACTH intersinus gradient of ≥1.4 at any time point after stimulation.7 The BIPSS procedural details and results for each case as well as any available medical, operative, and pathology reports were reviewed. Of note, we routinely administer heparin to patients intraprocedurally with dosage based on weight, on average giving 4000 units intravenously.

Figure 1

Image from a representative inferior petrosal sinus sampling (IPSS) case in a patient with a pituitary adenoma. IPSS venography after proper placement of both catheters (black arrows); star indicates the expected location of the pituitary gland within the cavernous sinus.

We considered CD to be confirmed based on BIPSS concordance with the histology of the surgical specimen or normalization of urine free cortisol levels after trans-sphenoidal surgical resection. Patients who did not meet the criteria of CD based on BIPSS were considered to have ectopic ACTH secretion; for these patients, identification of an extrapituitary source of ACTH was noted.

Tumor lateralization was determined in all patients with histologically confirmed CD. Tumor lateralization indicated by BIPSS was considered concordant if matched by intraoperative findings. Procedural complications were evaluated from the medical chart; assessment of venous thromboembolism was done clinically.

Finally, to compare the BIPSS results between patients who received DDAVP and those who received CRH (Acthrel; Ferring Pharmaceuticals, Parsippany, New Jersey, USA), we included all patients who received DDAVP and had centralizing and lateralizing BIPSS results. We included an equivalent number of patients who received CRH in the consecutive months preceding the change to DDAVP, excluding those with non-centralizing non-lateralizing BIPSS. We then used the side to which BIPSS was lateralized and calculated the ratio of the ACTH level on that side to the ACTH level in the periphery at the same time point. Since two baseline measurements were taken for each patient, these measurements were averaged. We then compared the DDAVP group and CRH group with two-way ANOVA to assess whether there was a significant difference in these lateralized ratios.


A total of 20 patients (table 1) underwent BIPSS with DDAVP stimulation (4 men and 16 women, mean age 38 years, range 12–84). Weight gain (n=15), skin changes (n=15), and facial rounding (n=13) were the most commonly reported symptoms. Sampling was successful following bilateral catheter positioning in all 20 patients. No complications were observed during or immediately following the procedures. In particular, no complications specific to DDAVP including venous thrombosis were encountered.

Table 1

Demographic, clinical, and imaging features of patient population. Data are presented with findings at BIPSS and surgery

Of the 20 patients, 18 had centralizing BIPSS suggestive of CD. One of the two patients who had a non-centralizing BIPSS was lost to follow-up and the other patient underwent trans-sphenoidal surgery and was found to have a pituitary adenoma (false negative). This patient had a urine free cortisol level of 116.7 μg/24 h (normal 17–47) and a late night salivary cortisol of 9 nmol/L (normal <4.3) at the time of the procedure. Furthermore, catheter positions appeared appropriate from BIPSS venography. Since the other patient was lost to follow-up, the specificity of BIPSS with DDAVP could not be calculated despite this one false negative case.

Of the 18 patients with centralizing BIPSS, 17 underwent trans-sphenoidal surgery and all were found to have pituitary adenomas. One patient had a positive BIPSS considered to be non-diagnostic as neither the free nor late night salivary cortisol was assessed at the time of BIPSS. Periprocedural hypercortisolemia must be present for the accurate interpretation of BIPSS results as there are cases of episodic hypercortisolemia.7 For the sensitivity analysis we excluded this patient who did not receive surgical intervention. Because of the one false negative case, the calculated sensitivity for BIPSS with DDAVP stimulation is 94.4%.

We evaluated concordance of tumor lateralization in the 15 patients who had both lateralizing BIPSS results and reported lateralization in the surgical note (table 1). In these 15 patients, lateralization was concordant in 10 patients (66%).

Sixteen of the 20 patients had centralizing and lateralizing BIPSS results. We also selected 16 patients who underwent BIPSS with CRH who had centralizing and lateralizing results in the consecutive months preceding the shift to using DDAVP. Two-way ANOVA showed no significant difference between DDAVP and CRH stimulation in the ACTH ratios of the lateralized side (F=0.04; p=0.85). There was a significant effect of time on the ACTH ratio (F=15.9; p<0.0001), with the values tending to increase to a maximum immediately after stimulation and then decreasing slowly over time (figure 2).

Figure 2

Similarity of bilateral inferior petrosal sinus sampling (BIPSS) results with corticotropic-releasing hormone (CRH) or desmopressin (DDAVP) stimulation in a matched number of patients with centralizing lateralizing BIPSS. The mean ratios of the adrenocorticotropic hormone (ACTH) level on the lateralized side relative to the peripheral ACTH level were calculated at each time point relative to CRH or DDAVP stimulation. No significant difference was observed between the two groups. There was a significant effect of time on the ACTH ratio (F=15.9; p<0.0001), with maximal values immediately after stimulation followed by a subsequent decrease.


Timely diagnosis and curative intervention of CD are imperative to limit its long-term morbidity. Localization of the source of ACTH secretion to distinguish CD from ectopic ACTH secretion initially involves non-invasive tests including the CRH stimulation test, the high-dose dexamethasone suppression test, and MRI of the pituitary.1 ,8–10

The current shortage of CRH in the USA precludes its use as either a non-invasive laboratory test or in BIPSS. Even when available, however, the non-invasive CRH stimulation test and the dexamethasone suppression test have less than optimal sensitivity and specificity for diagnosing CD, ranging from 50% to 70%.1 MRI of the pituitary is limited as the vast majority of ACTH-secreting pituitary tumors are sub-centimeter microadenomas detected in only about 60% of cases, and because of the high prevalence (up to 10%) of non-functioning incidentally discovered pituitary lesions in the general population resulting in false positive scans.11 BIPSS therefore remains the gold standard test to diagnose CD with an extremely high diagnostic accuracy to justify the decision to perform invasive surgical resection.

BIPSS is associated with a very low (0.2%) incidence of serious complications when performed by an experienced operator. However, thromboembolic complications have been reported after BIPSS in the absence of prophylactic anticoagulation.4 ,5 ,12 Although previously reported events were unrelated to DDAVP use, theoretical concerns exist for DDAVP as it increases the levels of vWF and factor VIII13 and could theoretically act as a procoagulant agent. As CS is characterized by a hypercoagulable state, we routinely heparinize all patients undergoing BIPPS. Furthermore, despite these concerns, DDAVP is readily indicated for routine administration at higher repeat doses in the treatment of diabetes insipidus. The risk of adverse effects is expected to be much lower with a single administration (as in BIPSS), and our study corroborates its safety as previously demonstrated.14

Studies have shown a sensitivity and specificity exceeding 90% for BIPSS with CRH stimulation.2 Due to the low incidence of CD, our series of BIPSS with DDAVP stimulation was limited to 20 patients. While previous studies have shown good accuracy for BIPSS with DDAVP,14 our series is the first to directly compare BIPSS with DDAVP and with CRH at the same institution. The results show similar diagnostic accuracy and complication rates, suggesting that DDAVP is a safe and effective alternative to CRH with a sensitivity of approximately 94.4%. Specificity could not be calculated because of the two negative BIPSS results. One patient was lost to follow-up and the other patient was a true false negative despite suggestive urinary free cortisol and late night salivary cortisol findings consistent with hypercortisolemia; it is not clear why this patient's result was negative.

Despite its impressive diagnostic accuracy, BIPSS with either CRH or DDAVP stimulation proved to be an unreliable test (66% accuracy) for lateralization of a pituitary adenoma. This finding is in concordance with previous studies15 and is probably due to anatomical variation in venous drainage.7

Also of note is the wide availability and cost-effectiveness of DDAVP compared with CRH. CRH is not available in certain countries and is significantly more costly than DDAVP. For example, at our institution 10 µg DDAVP used in one procedure costs approximately $46.00 while 100 µg CRH commonly required for the same procedure costs more than 15 times as much. With accumulating evidence that CRH offers no apparent diagnostic advantage over DDAVP, the possibility of substituting CRH with DDAVP should be considered in view of rising healthcare costs.

In conclusion, our series comparing DDAVP with CRH for stimulation in BIPSS shows that DDAVP is a safe, sensitive, cost-effective and practical alternative to establish the diagnosis of CD.



  • Contributors Conception and design: RO. Analysis and interpretation of data: ARD, BA, RO. Drafting the article: ARD, JR. Revising it critically for important intellectual content: RO, JAH. Final approval of the version to be published: all authors.

  • Competing interests None.

  • Ethics approval Ethics approval was obtained from the Institutional Review Board.

  • Provenance and peer review Not commissioned; externally peer reviewed.